Australian Insurance Accounting
– A Prelude to IFRS Phase II ?
Tony Coleman
Chief Risk Officer IAG
Institute of Actuaries of Australia
Reserving for General Insurance Seminar
2
Agenda
•
•
•
•
•
An Introduction – IFRS for Insurance
Related developments in IFRS
Other Issues on the Horizon
Where we are now ?
Impact for Australian Non-Life Insurers ?
3
Insurance – Phase II
• Original Insurance Project started in 1997
• DSOP presented in 2001
• Project couldn’t complete by 2005 → 2 phases
– IFRS changes too much for a “single bite”
– Some areas of change needed more clarification and
debate (e.g. FASB viewpoint, fair value principle)
• Phase I – completed by issue of IFRS 4 in March 2004
and implemented (in Aust) through AASB changes from 1
Jan 2005
• Phase II – re-ignited in late 2004 (Std: 2009+ ?)
• Influence of parallel projects
4
5
User Needs Survey
• Survey used by CFO Forum to assess users’
views on key needs
• Identify priorities of users including investors,
analysts, hedge funds and rating agencies
• 50 responses from 220 invitees covering Europe,
USA, Japan and Australia
6
7
8
9
10
11
Conclusions from Survey
•
There is a significant mismatch between the actual and preferred
quality of financial statements
•
There is also a significant mismatch between the views of users and
preparers on this issue
•
More detailed disclosure is needed of information relevant to users
•
Users are generally concerned with understanding the key drivers
which will allow them to improve forecasts for the future performance
of the business
•
Reliability of information should be improved and made less subject
to subjective management decisions
12
Source: Lehman Brothers
13
Source: Lehman Brothers
14
IASB Concepts & other Projects
Related to Phase II
• Accounting Framework changes
• The philosophy of “Fair Value”
• Financial Instruments – IAS39
• Revenue recognition – will insurance be the
“tail wagging the dog” ?
• US GAAP and IASB convergence
15
Phase II – The Accounting
Framework
• The outcome needs to reflect the Accounting
Framework – i.e. allow for transparent,
prospective view of measurement of risk
• Consistency of accounting treatment for
assets and liabilities vs. potential impact of
IAS 39
16
Conceptual Accounting Frameworks
•
FASB (in “Concepts Statement No 1”) :
“…Financial reporting should provide information to help investors,
creditors, and others assess the amount, timing and uncertainty of
prospective net cash inflows to the related enterprise” (para 39)
•
The IASB framework:
“The economic decisions that are taken by users of financial
statements require an evaluation of the ability of an entity to
generate cash and cash equivalents and of the timing and certainty
of their generation” (para 15)
17
Potential to manipulate
• Is malleability a problem?
–
–
–
•
Reliable numbers mean investors can make sound decisions
Bad numbers lead to loss of confidence in markets and shareholder loss
Concerns on “hidden” liabilities e.g. environmental (carbon / water etc)0
Potential to manipulate increases
– As businesses become more complex and more “knowledge” based
– Move to “fair value” accounting
Study from Harvard and Chicago University found ample evidence of tinkering at
delicate moments before acquisitions, equity offerings and exercising stock options
e.g. inflated assumed rate of return on pension fund assets thus flattering profits,
health care liabilities, stock options not expensed, etc
18
IFRS – Impacts on Corporates
Traditional accounting
Fair value accounting
Largely static and objective
But often irrelevant historical costs
e.g. derivatives, options, swaps
Updates items regularly in order
To try and capture current value
But more malleable
•
In June 2005, America’s SEC endorsed shift to fair value accounting which it
thinks will
– lead to more realistic accounts and
– reduce firms interest in structuring transactions to meet accounting goals
•
FASB release in late 2005 on guidance on how to apply fair value devising
“hierarchy” of items based on hardness to value
19
“Fair” Value - Issues and Risks
• More volatile profits
- Any change in Company’s Balance Sheet is taken into Profit & Loss A/c
• More reliance on estimates
- Many items are not traded in liquid markets & lack verifiable price
- Determining “fair value” involves complicated models based on subjective estimates
- At best estimates are difficult for investors/auditors to verify or compare across
companies
- At worst unscrupulous could manipulate models to flatter bottom line e.g. Enron
For example, research shows fair value of bank loans can vary widely depending on inputs/
methodologies. Market values for lower-rated corporate bonds, one possible benchmark,
can vary by as much as 2 – 5% giving managers leeway to fiddle with numbers
Source: Economist: Crooks and Books 28.7.2005
20
Fair Value Solutions & Consequences
Potential solutions
•
•
•
•
•
•
•
Similar issues
that actuaries
face
Disclose assumptions behind estimates
Disclose how different assumptions would affect profits
Reveal whether estimated values have altered over time & why
For key items show historical cost & updated fair value explaining derivation
Objectivity and independence of those making estimates and those verifying them
Harder scrutinising of numbers by investors
Better disclosure e.g. how much profits are based on facts compared to estimates, how
previous estimates panned out
Consequences
•
•
•
Encourage managers to value things realistically
Highlights difficulty of measuring a firm’s performance – a complex subjective notion which
cannot be captured in a single profit figure
Helps investors decide whether estimates are fair or foolhardy and to think harder about
what a business is worth
21
So - What issues does this raise ?
•
When can we measure “Fair Value” reliably ?
•
What do we do when we can’t ?
•
Are we monitoring and reporting risk issues adequately ?
•
How are others dealing with problems we also have to deal with
e.g. estimates, assumptions, judgements ?
•
What impact will this form of reporting have on investment decisions ?
•
What is the relationship between profits, risk and share value ?
•
Are the rules resulting in appropriate outcomes and behaviours ?
•
Are we considering all of the issues that we should be ?
22
Phase II – Fair Value Issues
• “Full” fair value or compromise approach (based on latest IASB
proposal)?
• Credit risk adjustments (discount rate)
• Disclosure requirements
• Source of “market value” risk margins – entry vs. exit calibration
• Reliability of assessment
23
Phase II – Risk Adjustment
• Non-Life vs. Life
– Progression – simple to complex
• Stochastic approach issues:
– Choice of PoS/PoA vs “Cost of Capital” approach
– Unit of Account ?
– Diversification effects
• Deterministic approach issues:
– Lack of transparency
– Lack of consistency
24
Phase II – Revenue and Contract
Liability Recognition
Consistency in General Recognition Issues – e.g. identification of
legal obligations and “constructive” obligations, recognising that legal
and constructive definitions vary by jurisdiction
Other Detailed Recognition Issues
– DAC identification, differences by life/non-life etc (may encompass
“Origination” costs rather than just “Acquisition” costs)
Substantial Cultural Differences
regarding various contractual forms based on civil law (Napoleonic) vs.
common law (Anglo-Saxon) origins
25
Phase II –Issues Specific to Life
Insurance
• Policyholder Options and Behaviour
– Including allowance for renewal premiums, other continuation
issues and minimum deposit floor
• Performance Linkage
– Non-guaranteed elements, discretionary par features (timing only
or not), legal context etc.
• Participating Business
– IASB debating – when is it equity vs liability ?
– Answer appears to be “Equity” with some exceptions
– IAA completing a survey of types of par business around the world
26
Phase II – FASB and IASB “Modified
Joint Approach”
• Background: Major differences in US GAAP and IFRS
principles, which created opposition to change by FASB
• Solution: Discussion paper initiated by IASB to enable
convergence in approaches
• Current progress: Slow – FASB has recognised
difficulties caused by not splitting service from risk
(because US GAAP includes no definition of “contract”)
27
Phase II – Pressure on US GAAP
• Recent focus by US regulators on finite
reinsurance
– Contracts regarded as “bread and butter” (in the US) now being
challenged for existence of risk transfer (e.g. General Re and AIG)
• SEC identification of the need for IFRS
recognition
– EU/SEC agreement on a “roadmap” for elimination of the need for
companies using IFRS to reconcile with US GAAP by as early as
2007
28
Phase II – The Australian Situation
•
Phase I represented a positive step in the liability area
e.g. Liability Adequacy Test, mandatory Risk Margins
•
Australian approach to many Phase II issues appears to be finding
favour with the IASB e.g. prospective accounting approach in
general, using discounting and risk margins with appropriate
disclosure - all supported by IASB conclusion for non-life business in
May 2005. (No other jurisdiction uses this combination at present)
•
Recent IASB decision appears to favour elimination of Unearned
Premium and DAC concepts in favour of fully prospective approach
(similar to APRA premium liabilities for non-life)
29
Issues on the Horizon
• IAIS – International convergence on regulation (incl
capital requirements)
– Leveraging off IFRS + globalisation
• Review of IASB Framework ?
– “probable” / “likely” → “probability weighted”??
– “stand ready” obligations
– IAS 37
• Insurance “risk margins” – how ?
30
Risk Margins
• Liability must be based on unbiased estimate of the
mean of expected outcomes
• Risk margin calibration will not be by simple use of
Probability of Sufficiency
• Use of “Cost of Capital” concept likely
31
Linking PoS With “Fair (Market)
Value”
• Compared:
– NPV of the Cost of Capital required to hold capital in excess of the central
estimate of liabilities up to 99.5th percentile of outstanding claims
distribution throughout the claims run-off, discounted at a risk free rate
– Equivalent to initial capital needed less NPV of expected capital released
as claims are settled over time
• Assumed:
– Claims log-normally distributed (i.e. skewed outcomes)
– CoV’s constant over run-off of claims
– Realistic returns required on capital
• Tested:
– Short, “Medium” and Long Tailed classes
– 1 year, 3 years and 10 years average term to run-off claims respectively
32
Linking PoS With “Fair (Market)
Value”
Key assumptions (all considered realistic):
• CoVs used 10%, 20% and 30% for short, medium and
long tail portfolios respectively
• Risk Free rate 5% p.a.
• Cost of Capital 9% p.a. & 11% p.a. (i.e. 4% p.a. and
6% p.a. over risk free rate)
• Examined various diversification effects between
combinations of portfolios
33
Linking PoS With Market Value Conclusions
• For classes modelled, the percentile of the liability
representing a realistic result varies significantly by
duration:
• Around 55%-60% PoS for short tail
• Around 65%-75% PoS for medium tail
• Around 80%-90% PoS for long tail
• PoS equivalent of “Fair (Market) Value” varies according
to the key risk attributes of a portfolio
34
Australian Insurance Accounting
– A Prelude to IFRS Phase II ?
Tony Coleman
Chief Risk Officer IAG
Institute of Actuaries of Australia
Reserving for General Insurance Seminar
35